The proposed research will provide important insight into two serious problems for veterans' healthcare: 1) mixed effectiveness of communication improvement interventions in the VA to date, and 2) persistence of racial/ethnic disparities in VA healthcare. We have identified 4 best acts of physician communication that patients believe to be most important in a clinical interaction. We propose to develop and evaluate a physician communication intervention that focuses specifically on these acts. In addition, our abundance of pilot study data from patients of various races/ethnicities enables us to create an intervention applicable to communication that spans patient race/ethnicity. Evaluating this intervention with physicians and patients in 2 independent VA primary care settings will help determine whether a patient-driven communication focus that incorporates input from patients of varied race/ethnicity can help close the gap in health outcomes between white and minority veterans. We aim to (1) demonstrate the efficacy of using best acts of physician communication as content for an intervention to improve physician communication skills overall and reduce variation in communication with white versus racial/ethnic minority patients, and (2) assess the impact of patient experiences of best acts on patients' perceptual and behavioral outcomes. We have collected data from over 200 primary care patients as they viewed a video of themselves and their physician interacting during a clinic visit. From these data, we are developing an educational intervention for physicians that focuses specifically on the 4 best acts we have identified. Scripts of the 4 best acts will be developed, and standardized patients at Baylor College of Medicine will perform the acts, which will be video recorded. We will then conduct a 2-site longitudinal observational study of 16 primary care physicians (8 at each site) and 36 patients of each physician. At the first site at 3-month intervals, physicians will attend an instruction session consisting of viewing a video of one best act and participating in group discussion, role-playing re-enactment of the act, and formulation of bulleted take-away points to facilitate subsequent recall. Once completed, the sequence of sessions will occur at the second site. At both sites, clinical interactions of participating physicians and patients will be audio recorded, and audio recordings will be analyzed for frequency of occurrence of the 4 best acts. After each study visit, each patient will complete validated surveys measuring patients' perceptions of their physician on 4 affective dimensions known to impact patients' subsequent perceptual and behavioral outcomes. Efficacy will be assessed across patient demographics.